Probing the reasons behind the changing pace of warming

Whatever happened to global warming? Right now, that question is a good way of starting a heated argument. Some say it is steaming ahead. But others say it has stalled, gone into reverse, or never happened at all - and they don't all run oil companies or vote Republican.

So what is going on?

First, talk of global cooling is palpable nonsense. This claim relies on the fact that no year has yet been hotter than 1998, an exceptional year with a huge planet-warming El Nino in the Pacific Ocean. Naysayers pretend that 1998 was typical, when it was anything but, and that temperatures have been declining since, which is statistical sleight of hand.

Meanwhile consider this. According to the National Oceanic and Atmospheric Administration (NOAA), all 12 years of the new century rank among the 14 warmest since worldwide record-keeping began in 1880. The second-warmest year on record, after 1998, was 2010. This is not evidence of cooling.

But there is a growing consensus among temperature watchers that the pace of warming in the atmosphere, which began in earnest in the 1970s and seemed to accelerate in the 1990s, has slackened, or stalled, or paused, or whatever word you choose. It may turn out to be a short blip; but it is real. "Although the first decade of the 21st century was the warmest on record, warming has not been as rapid since 2000," says Pete Stott, head of climate monitoring and attribution at the UK's Met Office, one of the leading keepers of the global temperature. He calls it a "hiatus" in warming.

In a blog last week, James Hansen, the retiring head of NASA's Goddard Institute for Space Studies (GISS), agreed that "the rate of global warming seems to be less this decade than it has been during the prior quarter century"

Something is going on. With heat-trapping greenhouse gases accumulating in the atmosphere ever faster, we might expect accelerated warming. So it needs explaining.

There are a number of theories. Hansen suggested that extra emissions of particles in Asian smogs could be shading the Earth and camouflaging the greenhouse effect. In a February post on RealClimate, his Goddard Institute colleague Gavin Schmidt instead pointed to fewer warming El Ninos and more cooling La Ninas. He suggested that adjusting for their influence produced an unbroken pattern of warming.

Schmidt's analysis certainly hints at a role for the oceans in all this. And most researchers on the case argue that, one way or another, the most likely explanation for the heating hiatus is that a greater proportion of the greenhouse warming has been diverted from the atmosphere into heating the oceans. A new study from Kevin Trenberth of the National Center for Atmospheric Research in Boulder, Colorado, published online inGeophysical Research Letters, found that ocean warming has been accelerating over the last 15 years.

Richard Allan of the University of Reading in England says simply: "Warming over the last decade has been hidden below the ocean surface." If you take the oceans into account, he says, "global warming has actually not slowed down."

This should not come as a surprise, notes Chris Rapley of University College London. The oceans are the planet's main heat sinks. More than 90 percent of the extra heat trapped in the atmosphere by greenhouse gases ends up there. But, while climate models are good at calculating atmospheric processes, they are poorer at plumbing the ocean-atmosphere interactions that determine how fast and how regularly this happens.

That makes those interactions a big source of uncertainty about atmospheric global warming, especially over the short term. If oceans grab a bit more heat one year, they can shut down that year's warming. Equally, if they release a bit more they can accelerate atmospheric warming. This matters. "The way the ocean distributes the extra energy trapped by rising greenhouse gases is critical... [to] global surface temperatures," says Allan. For forecasters trying to figure out the next decade or so, oceans could screw it all up.

Some bits of the puzzle have been known for a while. For instance, during El Nino years, warm water spreads out across the equatorial Pacific and the ocean releases heat into the air, warming the air measurably. That is what happened in 1998.

But while El Ninos come and go within a year or so, there are other cycles in heat distribution and circulation of the oceans that operate over decades. They include the Pacific Decadal Oscillation and the Atlantic Multidecadal Oscillation (AMO), both of which have been implicated in climate fluctuations in the 20th century. So have these or other ocean cycles been accelerating the uptake of heat by the oceans?

Virginie Guemas of the Catalan Institute of Climate Sciences in Barcelona, believes so. In a paper published in Nature Climate Change this week, she claims to provide the first "robust" evidence linking ocean uptake of heat directly to what she calls the recent "temperature plateau" in the atmosphere.

By plugging detailed measurements of recent atmospheric and sea temperatures into EC-Earth, a European model of interactions between atmosphere, oceans, ice and land surfaces, Guemas found that about 40 percent of the take-up was in the tropical Pacific, and another 40 percent in the tropical and North Atlantic.

She told me that it seems likely the changing thermohaline ocean circulation, which starts in the North Atlantic, plus the cycles of El Nino and perhaps the AMO, may play a prominent role. She thinks her model could have predicted the recent slowdown of atmospheric warming ahead of time.

That would be a breakthrough, but nobody has done it yet. Meanwhile, the climate modellers are skating on thin ice when they make predictions that play out over the timescales of a decade or so on which ocean cycles seem to operate. These forecasters can claim that, all things considered, they have done pretty well. But the forecasts remain hostages to fortune.

If anything, the recent pause shows the model forecasts in a good light. Myles Allen, a climate modeller at Oxford University in England, reported in Nature Geoscience last month on an audit of one of his own forecasts, which he made in 1999. He had predicted a warming of a quarter-degree Celsius between the decade that ended in 1996 and the decade that ended in 2012. He found that, in the real world, temperatures got too warm too soon during the 1990s; but the slackening pace since had brought the forecast right back on track.

That shows the forecast is performing well so far, but Allen admitted it might not stay that way. If temperatures flat-line out to 2016, his model's prediction for that year will look no better than a forecast based on a series of random fluctuations.

Some in the mainstream climate community privately admit that they were caught out by the slackening pace of warming in the past decade or so. Back in the 1990s, some suggested - or at least went along with - the idea that all the warming then was a result of greenhouse gases. They were slow to admit that other factors might also be at work, and later failed to acknowledge the slowdown in warming. As Allen pointed out earlier this year: "A lot of people were claiming in the run-up to the Copenhagen 2009 [climate] conference that warming was accelerating. What has happened since then has demonstrated that it is foolish to extrapolate short-term climate trends."

Not surprisingly they have been taken to task for this hubris. Roger Pielke Jr., an environmental studies professor at the University of Colorado at Boulder, who enjoys baiting the mainstream, told me last month: "It is good to see climate scientists catching up with the bloggers. They should ask why it took so long to acknowledge what has been apparent to most observers for some time."

But modellers are now responding more actively to the new real-world temperature data. For instance, the Met Office's Stott reported last monththat global temperatures were following the "lower ranges" of most model forecasts, and that higher projections were now "inconsistent" with the temperature record.

And last December, the Met Office downgraded its best guess for temperatures in the five years to 2017 from 0.54 degrees C higher than the average for the late-20th century average to 0.43 degrees higher. It said the new forecast was the first output of its latest climate model, HadGEM3, which incorporates new assessments of natural cycles.

But the problem is that these cycles are not well integrated into most climate models. Natural cycles could switch back to warming us again at any time, admits Stott. But he has no clear idea when.

The stakes for the climate forecasting community are high. It may be unfair, but the brutal truth is that if the climatologists get their forecasts for the coming decade badly wrong, then a great many in the public will simply not believe what they have to say about 2050 or 2100 – even though those forecasts may well be more reliable.

Forecasters badly need a way to forecast the ocean fluctuations, and it could just be that Guemas's new study will help them do that. She claims that her findings open the way to the future delivery of "operational decadal climate predictions." For now she is cautious about making firm predictions, but told me she believes that "the heat that has been absorbed recently by the ocean might very well be released back to the atmosphere soon. This would be the scenario of highest probability. It would mean an increased rate of [atmospheric] warming in the next decade."

It would indeed. If natural cycles start pushing towards strong warming, they will add to the continued inexorable upward push from rising concentrations of heat-trapping greenhouse gases. In that case, we would see climate change returning to the rapid pace of the 1990s. Whatever happened to global warming? The odds may be that by 2020 it will have come roaring back.